[fitswcs] Polarization codes

[Doug Tody]

Hi Mark -

You are right, flame-bait was probably too strong a term (it had been
a long day).  I don't think there is any point in us getting into a
long discussion of the history of WCS; I was merely objecting to your
suggestion that IRAF merely adopted WCS after it had been developed,
implemented, and proven in AIPS.  That was too much to let pass
without a comment.  But I don't want to let this distract us from
the very important polarization discussions.

> Original versions of G&C (it started as a single paper) used the CDi_j
> variant of the CD00i00j matrix proposed by Hanisch & Wells (1988) which
> came from the NASA-sponsored meeting held in Charlotesville in Jan/1988.
> Isn't that where IRAF got the CD matrix?

Indeed this is where the CD matrix came from - it grew out of work
done primarily by the STSDAS group at STScI.  Bob and the STSDAS
group at the time were members of the overall IRAF project, of course.
The CDi_j notation is what is used in IRAF MWCS (see below), although
I don't remember anymore exactly how we came up with this notation.

> > - the current WCS implementation
> >in IRAF predates the WCS standard.
> 
> Doesn't it use WCSLIB?  Lindsay provided me with much valuable feedback
> on the earlier versions of WCSLIB.  I was given to understand that she
> was modifying IRAF to use it.  (Sadly, AIPS, Miriad and karma also have
> not progressed beyond the AIPS convention and I'm sure there are many
> more.)

No, IRAF does not use WCSLIB, that came along much later.  The IRAF
implementation of WCS is MWCS, which was introduced in 1989.
Many modern WCS concepts were introduced in MWCS, although they may
well have been developed independently in AIPS at about the same time.

Cheers,

	- Doug